Previously, there was no way to specify which site you wanted to import the entries to. Imports appeared to succeed, but you could not get the data to display.

A site_id parameter has been added to the plugins to allow you to select the id of the weblog into which you want the data imported.

Current versions of FeedGrab and CSVGrab can be downloaded from the Products page.

brandnewbox is becoming a more business orientated site, with more emphasis in what has become my main source of work - ExpressionEngine. This site in the coming months will include more ExpressionEngine and web development articles and tutorials, and I will start adding more work to the Products section the site.

Please bear with me while I update the site, and let me know if you have any comments.

As you have no doubt heard by now, the Safari 3 Public Beta is now available for OS X and (surprise!) for Windows XP and Vista.

From: Surfin’ Safari

I find Safari a great little web browser (even if Firefox with Firebug trump it for web development), so this is great news. The beta runs without any trouble on my work PC running Windows 2000 (despite it being advertised as XP or Vista only) and it is quick and good to look at.

Get the Debug menu running on Windows by editing (in my case):

C:\Documents and Settings\aweaver\Application Data\Apple Computer\Safari\Preferences.plist

and adding:

<key>IncludeDebugMenu</key>
<true //>

No sign of the Web Inspector on Windows yet, though.

In the last 12 months all three have begun major redesigns. 

The Telegraph was the first to show their hand last May, followed by The Guardian’s Travel site later in the year. Yesterday’s launch by The Times completed the set and it is striking to see the similarities that the 3 groups have - almost certainly accidentally - achieved.

All have gone for a larger page width, with strong use of whitespace and grids. A prominent lead image focuses the attention on the main article and the use of advertising seems to have been reduced (or at least its appearance is less obvious). Content-wise there seems to have been a push to provide parts of the site to encourage users to contribute, and the prevalence of multimedia has increased.

All 3 designs are atrractive and effective, and with their white, strongly gridded design owe a nod towards their paper-based heritage. One interesting thing to see how that these 3 UK online news sources have taken their designs away from the style of the number one source, the BBC.

This article describes how to add an Opensearch function to your Expression Engine site using EE’s built-in search templates.

The Problem

Although not essential to the Opensearch syntax, IE7 is unable to handle searches that send their queries as POST requests. This is, unfortunately, how EE searches work.

The pMachine Knowledge Base provides a cunning AutoSearch answer, but as it requires Javascript will not always be a suitable solution.

The Autosearch plugin was developed to work around this problem. By creating an extra template and pointing the plugin to it, it is possible to send search requests on the URL (ie, as a GET request).

Setting Up

Firstly, download and install the autosearch plugin.

The system requires 3 templates.

The search results template

Your existing search results page will do for this. I’ll call this template search/results

The search box template

This should contain the code to generate the search box and ideally NOTHING ELSE. Don’t put any headers or text in this template - this template will never be seen by the user and is just used to set up the search. I’ll call this search/searchbox

{assign_variable:my_weblog="default_site"}
{assign_variable:my_template_group="site"}
{exp:search:simple_form weblog="{my_weblog}" result_page="search/results"}
<label for="keywords">Search:</label>
<input type="text" class="text" name="keywords" id="keywords" value="" size="18" maxlength="100" />
<input type="submit" class="button" value="Search" />
{/exp:search:simple_form}

The opensearch template

This is the template that initiates the search and should contain the call to the autosearch plugin.

{ exp:autosearch 
        url="http://www.brandnewbox.co.uk/search/searchbox" 
        keywords="{segment_3}" }

Again, the user will never see this page so don’t bother adding headers or footers.

Call this template (something like) search/opensearch

Testing the search works

Call the search from your browser’s address bar by calling the opensearch template followed by the search term, eg, http://www.your-site.com/index.php/search/opensearch/SearchTerm

Your search results page should be returned.

Autodiscovery of your Opensearch

First you need an Opensearch description file. I’ve saved this:

<OpenSearchDescription xmlns="http://a9.com/-/spec/opensearch/1.1/" xmlns:moz="http://www.mozilla.org/2006/browser/search/">
	<ShortName>brandnewbox.co.uk</ShortName>
	<Description>Search brandnewbox.co.uk</Description>
	<InputEncoding>utf-8</InputEncoding>
	<Url type="application/rss+xml" template="http://www.brandnewbox.co.uk/search/opensearch/{searchTerms}"/>
	<Url type="text/html" template="http://www.brandnewbox.co.uk/search/opensearch/{searchTerms}"/>
	<moz:SearchForm>http://www.brandnewbox.co.uk/search/
</OpenSearchDescription>

to:

http://www.brandnewbox.co.uk/opensearchdescription.xml

Then in your web pages’ header add:

<link rel=
"search" type="application/opensearchdescription+xml" title=
"Search brandnewbox.co.uk" href="/opensearchdescription.xml">

Then, on viewing your website in FF2 or IE7 you should see your Opensearch appear in the search box menu.

This first article implements a simple search box and results page.

Create the search box

This is a simple variation of the Simple Search Form from the EE documentation. This could be put into its own template or more likely be part of a global header embedded into every page.

{assign_variable:my_weblog="default_site"}
{assign_variable:my_template_group="site"}

{!-- your page header --}

{exp:search:simple_form weblog="{my_weblog}"}


Search:




 



{/exp:search:simple_form}

{!-- your page footer --}

An optional result_page="search/results” can be added to the exp:search:simple_form tag to specify the search results page.

An example search box can be seen here.

The search results page

This will be in its own template (eg, search/results) wrapped in the site’s header and footer.

For usability we could include the search box again (with the search term filled in):

{!-- your page header --}

{exp:search:simple_form weblog="{my_weblog}"}


Search:




 



{/exp:search:simple_form}

Before displaying the results:

<p>Searching for: <b>{exp:search:keywords}</b> found {exp:search:total_results} results.</p>

{exp:search:search_results}


{title}
{excerpt}


{/exp:search:search_results}


{!-- your page header --}

This will display the search results in a simple, definition list.

The main reasoning seems to be that because Outlook uses the Word engine to create emails, it should use Word to display them in order to get consistent results. However this does not account for the many emails sent from non-Outlook clients.

Surely, Microsoft are approaching the problem from the wrong end - shouldn’t Word produce HTML compatible with its IE rendering engine (and all other web browsers on other platforms).

With Outlook having ~75% corporate market share, is it time to go back to using tables and images?

More coverage at Campaign Monitor.

Clarke introduces techniques to harness CSS to create attractive and powerful sites, without forgetting accessibility, pointing out its often ignored relationship with usability.

The book looks beautiful, and reads easily. Sidenotes provide added depth when required but don’t interrupt the flow. A ream of yellow post-it notes are already sticking out of my copy, marking techniques and references I want to return to.

This book is full of inspirational ideas on all levels.

Usage is similar to FeedGrab, the template:

{exp:csvgrab url="http://www.url.co.uk/products.csv"
    weblog="5"
    title="1"
    skip="6"
    delimiter="TAB"
    encloser=""
    use="2|3"
    fields="prod_description|prod_keywords" }

imports a tab delimited CSV file into the weblog with id of 5, skipping the first 6 lines.

It uses column 1 as the title field, and columns 2 and 3 as the custom fields prod_description and prod_keywords.

delimiter is the character that separates the columns, encloser is the character that wraps the column.

eg, “Title”,"Description”,"Keywords" has a comma delimiter and a quote encloser.

EE ‘cleans up’ plugin parameters, so for delimiter and encloser use the following codes rather than the actual character: TAB, SPACE, QUOTE, eg,

delimiter="TAB" encloser="QUOTE" 

A forum thread exists on ExpressionEngine for bug reports, feature requests etc.

Download: EE CSVGrab plugin.

Introduction

The FeedGrab plugin allows you to insert the contents of RSS or ATOM feeds into Expression Engine weblogs.

Why?

The FeedGrab plugin, as well as allowing you to keep your own copy of the data in your own database, provides you with the ability to use all of Expression Engine’s nifty weblog features, such as calendar-based archives, the facility to randomly display items, search tools and the option of relating this external data with your own weblogs entries via EE’s related field feature.

With the data stored in a weblog you have the option of displaying the information however you like, either combined with other weblog data or on its own.

Install plugin

Download the latest version of the FeedGrab plugin from here.

To install it, follow ExpressionEngine’s plugin installation instructions

A look at the feed

The RSS feed I will be using is from mag.nolia.com and is a list of my most recently saved bookmarks. From the site I find its URL:

http://ma.gnolia.com/rss/lite/people/ajweaver

Loading this file in the browser and viewing its source shows its structure - the bits I’m interested in is between the <item> tags

    <item>
      <title>Compose to a Vertical Rhythm</title>
      <author>ajweaver</author>
      <description type="html"><p>Leading (line-height) a forgotten element of web typography.</p>
</description>
      <pubDate>2006-12-12T13:41:14-08:00</pubDate>
      <link>http://24ways.org/2006/compose-to-a-vertical-rhythm</link>
      <guid>http://ma.gnolia.com/people/ajweaver/bookmarks/mijolisc</guid>
      <category domain="http://ma.gnolia.com/tags">typography</category>
    </item>

For the purpose of this tutorial I want the title, description and link.

Create a new weblog

Before creating the weblog I usually set up the custom fields

Create a new weblog field group and add 2 new fields for description (make it a textarea) and link (a text input). I’ll call these magnolia_body and magnolia_url (Note, a weblog always has a title field by default so we’ll use that for the title field)

Create a weblog called magnolia and assign the custom fields group to it.

Tip: Edit the preferences for this weblog ( Weblog Administration > Weblog Management > Edit Preferences) and in Weblog Posting Preferences set Automatically turn URLs and email addresses into links? to No.

From the Weblog Management page make a note of the weblog’s id.

Create template

We’ll create a template to view the results (you might want to manually enter some test data to check this).

Create a new template group. It should have an index template automatically into which add:

{exp:weblog:entries
    weblog="events" orderby="date"
    limit="10" show_future_entries="yes"}

<h4><a href="{magnolia_url}">{title}</a></h4>
<p>{magnolia_body}</p>

{/exp:weblog:entries} 

Set up plugin

Create another template called go and add the call the plugin:

{exp:feedgrab url="http://ma.gnolia.com/rss/lite/people/ajweaver" 
                          weblog="WEBLOG_ID" 
                          title="title"
                          date="pubDate"
                          use="link|description" 
                          fields="magnolia_url|magnolia_body" }

Replace WEBLOG_ID with the value you made a note of earlier.

The title="title" line tells the plugin to use the feed’s title as the weblog entry’s title, and date="pubDate" tells it to use the feed’s pubDate field as the weblog entry’s publish date.

The use="link|description" fields="magnolia_url|magnolia_body" section tells the plugin to put link into magnolia_url and description into magnolia_body.

To call the plugin, simply go to http://www.yoursite.com/index.php/feedgrab/go

This should add all the new RSS entries into your new weblog. Subsequent calls will only add new entries.